DNA methylation is critical for essential processes including genomic imprinting, X-inactivation, transposable element silencing, and genome stability. DNA methylation patterns are the result of active DNA methylation and demethylation processes. Arabidopsis represents an excellent eukaryotic model system to dissect the functions and mechanisms of methylation and demethylation because of its facile genetics and genomics and the conservation of these processes with mammals. Arabidopsis has been a continued source of novel discoveries in the field of epigenetics. The function and regulation of DNA demethylation is under studied compared to processes that promote DNA methylation, partly because the mechanisms of active DNA demethylation have been elucidated only relatively recently. In plants, DNA demethylation is mediated by 5- methylcytosine (5-mC) DNA glycosylases that remove 5-mC from DNA by base excision repair, leading to its replacement with unmodified cytosine. Expression of the 5-mC DNA glycosylase ROS1 is significantly reduced in response to mutations in components of the maintenance methylation and RNA-directed DNA methylation (RdDM) pathways. This suggests that there is a molecular crosstalk between the methylation and demethylation pathways to achieve a cellular balance of these antagonistic enzymatic activities. The long-term goal of this research is to define the molecular mechanisms that ensure DNA methylation homeostasis during plant growth and development and to determine the functional consequences of perturbing that homeostasis. The proposed research focuses on ROS1, a locus where multiple epigenetic pathways converge to modulate DNA demethylase expression in response to global and local changes in DNA methylation. We will determine why and how ROS1 transcription is reduced in response to the disruption of the RdDM pathway. We will identify regulators of this process through a novel genetic screen that will identify genes that promote expression of ROS1 in a methylation deficient background. Finally, we will determine whether the decreased expression of ROS1 that occurs during wild type sperm development in response to reduced DNA methylation is responsible for establishing proper genome-wide DNA methylation patterns after fertilization. These efforts will significantly increase our understanding of how the activity of epigenetic pathways is modulated in response to the state of the epigenome. Our findings will be directly applicable to mammals and other eukaryotes that control their genome through DNA methylation based mechanisms.